Scaling TensorFlow and Caffe to 256 GPUs – NVIDIA Technical Blog News and tutorials for developers, data scientists, and IT admins 2025-07-23T19:02:16Z http://www.open-lab.net/blog/feed/ Brad Nemire <![CDATA[Scaling TensorFlow and Caffe to 256 GPUs]]> https://news.www.open-lab.net/?p=8938 2022-08-21T23:44:00Z 2017-08-08T17:43:01Z IBM Research unveiled a ��Distributed Deep Learning�� (DDL) library that enables cuDNN-accelerated deep learning frameworks like TensorFlow, Caffe, Torch and...]]> IBM Research unveiled a ��Distributed Deep Learning�� (DDL) library that enables cuDNN-accelerated deep learning frameworks like TensorFlow, Caffe, Torch and...

IBM Research unveiled a ��Distributed Deep Learning�� (DDL) library that enables cuDNN-accelerated deep learning frameworks like TensorFlow, Caffe, Torch and Chainer to scale to tens of IBM servers leveraging hundreds of GPUs. ��With the DDL library, it took us just 7 hours to train ImageNet-22K using ResNet-101 on 64 IBM Power Systems servers that have a total of 256 NVIDIA P100 GPU accelerators in��

Source

]]> 0 ���˳���97caoporen����